September 26, Word count: 7,542 (6,542 words + 4 figures) Submitted for presentation 94 nd Annual Meeting of the Transportation Research Board

Transcription

1 WHAT DO AMERICANS THINK ABOUT FEDERAL TAX OPTIONS TO SUPPORT PUBLIC TRANSIT, HIGHWAYS, AND LOCAL STREETS AND ROADS? RESULTS FROM YEAR 5 OF A NATIONAL SURVEY Asha Weinstein Agrawal Director, Mineta Transportation Institute s National Transportation Finance Center San José State University One Washington Square San Jose, CA asha.weinstein.agrawal@sjsu.edu phone: Hilary Nixon Associate Professor, Department of Urban and Regional Planning San José State University One Washington Square San Jose, CA hilary.nixon@sjsu.edu phone: September 26, 2014 Word count: 7,542 (6,542 words + 4 figures) Submitted for presentation 94 nd Annual Meeting of the Transportation Research Board

2 Agrawal & Nixon 2 ABSTRACT This paper summarizes the results of a national random-digit-dial public opinion poll that asked 1,503 respondents if they would support various tax options for raising federal transportation revenues, with a special focus on understanding support for increasing revenues for public transit. Eleven specific tax options tested were variations on raising the federal gas tax rate, creating a new mileage tax, and creating a new federal sales tax. Other questions probed various perceptions related to public transit, including knowledge and opinions about federal taxes to support transit. In addition, the survey collected data on standard socio-demographic factors, travel behavior (public transit usage, annual miles driven, and vehicle fuel efficiency), and attitudinal data about how respondents view the quality of their local transportation system and their priorities for government spending on transportation in their state. All of this information is used to assess support levels for the tax options among population subgroups. INTRODUCTION Over the past several decades, the transportation revenues available from state and federal gas taxes have fallen significantly, especially in terms of inflation-adjusted dollars per mile traveled. Between At the same time, the transportation system requires critical and expensive system upgrades. Among other needs, a large portion of the national highway system requires major rehabilitation, and there is a growing desire at all levels of government to substantially upgrade and expand infrastructure to support public transit, walking, and bicycling. This dilemma major desired system improvements but inadequate revenues available to fund them can be resolved in only two ways: either the nation must dramatically lower its goals for system preservation and enhancement, or new revenues must be raised. If the latter is to happen, legislators must be convinced that increasing taxes or fees is politically feasible. One portion of the political calculus that legislators make when deciding whether or not to raise new revenues is, of course, considering likely public support for or opposition to raising different kinds of taxes. This paper contributes to the understanding of current public sentiment about increasing transportation taxes by presenting the results of Year 5 of a telephone survey investigating public opinion about a variety of transportation tax options at the federal level. The specific taxes tested were 10 variations on raising the federal gas tax rate or creating a new mileage tax, as well as one option for creating a new federal sales tax. In addition, the survey collected standard sociodemographic data, some travel behavior data, and attitudinal data about how respondents view the quality of their local transportation system and their priorities for government spending on transportation in their state. All of this information was used to assess support levels for the tax options among different population subgroups. The survey questionnaire described the various tax proposals in only general terms, so the study results cannot be assumed to reflect support for any actual proposal put forward. Nevertheless, the results show likely patterns of support and, more important, the public s relative preferences among different tax options. In 2012 (the third year), questions were added to understand perceptions related to public transit, including knowledge and opinions about federal taxes to support transit. Also, new questions were added to explore respondents knowledge of whether different levels of government help to pay for transit, their opinion about whether gas tax revenues should be spent on transit, and their support for different Congressional options to raise additional revenues to improve and expand transit.

3 Agrawal & Nixon 3 Because the survey was the fifth year of a project to assess how public support for federal transportation taxes may change over time, most of the questions asked were identical to those in the earlier surveys carried out in the series (1, 2, 3, 4). This paper compares the results of the five surveys to establish how public views may have shifted over the past years. Readers interested in seeing a more extensive discussion of the study results may consult the project report on which this paper is based (5). EXISTING SURVEY RESEARCH To provide context for interpreting the survey results presented in this paper, this section reviews the results from 92 other public opinion polls that asked about support for gas, mileage, and sales taxes whose revenues would be used for transportation purposes. Details on the 92 polls are available in (5). The surveys, mostly conducted in the past eight years, were identified by searching the Internet-based archives of popular pollsters and aggregators of public opinion polls, including the Pew Center for the People and the Press, the Roper Center for Public Opinion Research, Rasmussen Reports, SurveyUSA, and PollingReport.com. This work was supplemented by searching Google to find mainstream media coverage on polls about transportation taxes. Complete survey results were obtained directly from the survey sponsors websites or through personal contact with the sponsors. Most of the surveys reviewed were conducted by public agencies, advocacy groups, popular pollsters, or news media; a few others were conducted by academics or research-oriented nonprofits. Gas Taxes Gas taxes are a primary source of transportation revenue at both the state and the federal level. However, the federal government and many states have not raised the tax rates in two decades, so the real value of the revenues collected has fallen with inflation. As a result, there is frequent talk about raising gas tax rates, and public opinion on such increases has been extensively polled. Making direct comparisons among the polls is difficult, because the specific tax increases proposed and the contexts in which they are presented both vary widely. For example, some proposals call for unspecified increases in the gas tax, while others propose specific increases that range from 5 to $2 per gallon. Some polls link the gas tax increase to a particular purpose, such as maintaining bridges, while others link the increase to very general uses, such as to help meet new transportation needs. Two general trends do emerge across the polls, however. First, support levels tend to be below 50% and are often considerably lower. Only about a quarter had support levels over 40%. Second, support tends to be particularly high when the tax increase is linked to some sort of environmental benefit. Table 20 in (5) presents the results for the 14 polls that link a gas tax increase with environmental benefits, shows that more than two-thirds of these found support levels above 40%. Mileage Taxes Far less polling has been done about mileage taxes because these are not currently in use anywhere in the United States, although they are under active discussion among transportation policymakers and researchers. A review of 17 polls shows that support is not especially strong but can be strengthened when the taxes are linked to environmental benefits (see Table 21 in (5)). The six polls linking a mileage tax to environmental benefits found support levels ranging

4 Agrawal & Nixon 4 from 33% to 50%, but the other eleven polls without that environmental link found support levels no higher than 39%. Sales Taxes Public opinion about local sales taxes to fund transportation programs has been extensively tested. However, very little polling has been done to test public support for a national sales tax to support transportation, most likely because the federal government does not collect sales taxes, leaving them for state and local governments to use as a revenue tool. (If the federal government were to consider imposing its own sales tax, there would likely be a very strong backlash from state and local officials.) For more than a decade, sales taxes have been one of the most popular methods used by local governments to raise revenue for transportation purposes. In almost all cases, the taxes were placed on the ballot for voter approval, so the election results provide one clear picture of the level of public support. And, in fact, many of these local sales taxes have passed, especially in California, where the great majority of the population currently lives in counties whose voters have approved local sales taxes for transportation by two-thirds majorities. In addition to the evidence from election results, considerable public polling has been done prior to elections to assess the appeal of sales tax increases. Table 22 in (5) summarizes a sampling of 30 polls testing public opinion on sales taxes. Overall support levels were quite high: 16 of the polls showed support at 50% or higher. SURVEY METHODOLOGY Questionnaire Design The survey questionnaire was designed to test public support for three types of taxes: an increase in the federal gas tax, a new national mileage tax, and a new national sales tax. In all cases, respondents were told that the revenue raised would be dedicated to transportation purposes. To make these hypothetical taxes easier for respondents to understand, the survey gave specific amounts for each. The amounts were selected to be simple numbers within the range of mainstream current policy discussion. Because a gas tax and a mileage tax are revenue options likely to receive considerable policy scrutiny in coming years, the survey tested support for these concepts when the taxes were presented in different forms. Overall, 11 different tax options were tested eight variants of a gas tax increase, two variants of a new mileage tax, and one new sales tax option. The gas tax was explored in the most depth because, at least for the near term, the most likely source of additional federal revenues would be an increase in the gas tax rate. All variants were included in all three years of the survey, except the gas tax questions about maintenance and safety, which were not included in the first year of the project. The specific tax options tested were as follows: Gas tax increases. Every variant of a gas tax increase involved raising the existing 18 per gallon tax to 28 per gallon, but each included a different set of information for respondents to consider. (The current federal tax on gasoline is 18.4 per gallon, but respondents were told that it was 18 per gallon to make the survey simpler to understand.) The eight variations were: A base-case 10 increase in the gas tax without further stipulations.

5 Agrawal & Nixon 5 A 10 increase in the gas tax that would be phased in over five years, increasing by 2 a year. A 10 increase in the gas tax, with the revenues to be spent only for projects to reduce local air pollution caused by the transportation system. A 10 increase in the gas tax, with the revenues to be spent only on projects to reduce the transportation system s contribution to global warming. A 10 increase in the gas tax, with the revenues to be spent only on projects to maintain streets, roads, and highways. A 10 increase in the gas tax, with the revenues to be spent only on projects to reduce accidents and improve safety. A 10 increase in the gas tax, with the revenues to be spent only on projects to add more modern, technologically advanced systems like real-time travel alerts, longer lasting pavements, and better timed traffic lights. A 10 increase in the gas tax, with respondents informed of the annual tax burden for a typical driver under both the current and increased tax rates. Respondents were told that the tax burden would increase from an average of $100 a year to $150 a year for someone driving 10,000 miles a year in a car with a fuel economy of 20 miles per gallon. New mileage taxes. Two variants of the mileage tax were presented, both of which involved levying a new tax per mile driven, with electronic meters being used to track miles driven and drivers being billed when they buy gas. The two variants, which differed only in the rate structure, were: A base-case 1 per mile tax, with every car being taxed at the same rate. A variable-rate mileage tax for which the average rate would be 1 per mile, but vehicles that pollute less would be charged less and vehicles that pollute more would be charged more. A new national sales tax. In this option, the federal government would levy a new 0.5% sales tax. The exact wording used to describe each tax to respondents can be found in Appendix A of (5), which reproduces the survey questionnaire. A new feature for the survey project introduced in the third year was a special focus on understanding support for raising revenues to support public transit. Questions were added asking respondents if they knew whether different entities help to pay for transit (transit riders, plus local, state, and the federal governments), their opinion about whether or not gas tax revenues should be spent on public transit, and their support for and preference among different Congressional options to find additional revenues to improve and expand transit services. For both support of the tax options and opinions about public transit, the survey was designed to assess how responses to the questions might vary by socio-demographic factors, travel behavior characteristics, and respondents opinions about their local and state transportation systems. Introductory questions asked respondents to rate the quality of roads and highways and transit service in their community and to indicate the priority they thought

6 Agrawal & Nixon 6 government should place on various options for improving the transportation system for everyone in their state. The questionnaire concluded with a standard set of socio-demographic questions on such factors as age, race and ethnicity, and income. To assess travel behavior, the survey included one question asking how many miles the respondent drove in the previous year and another question asking if the respondent had used any form of public transit within the previous 30 days. Respondents were also asked the average fuel efficiency of the vehicle they drove the most for personal travel. Survey Implementation The Social Science Survey Center at California State University, Fullerton, conducted the survey on behalf of the Mineta Transportation Institute s National Transportation Finance Center. The interviewing was completed in two phases, from March 4 April 7 and April 23 April 30, A total of 1,503 adults nationwide were interviewed by telephone in either English or Spanish, with 33 (2%) of the interviews conducted in Spanish. Telephone numbers included in this sample were randomly generated, and survey respondents were reached by both cell phone (N=352) and landline phone (N=1,151). Including both cell and landline phone numbers meets the best practices in current survey research (5, 6). According to a 2012 study by the Pew Center for the People and the Press, telephone surveys that include landlines and cell phones and are weighted to match the demographic composition of the population continue to provide accurate data on most political, social and economic measures (5, p. 1). Unless otherwise indicated, all results are weighted to match the Census Bureau s 2012 American Community Survey one-year estimates with respect to gender, race, Hispanic ethnicity, education level, annual household income, and age (7). The margin of error for the total sample is ± 2.53 percentage points at the 95% confidence level. Smaller subgroups have larger margins of error. SURVEY FINDINGS This section presents highlights of the survey results. It first describes the survey respondents and then presents support levels for the tax options among all respondents and also among population subgroups. The section concludes with findings on how support for the base-case 10 gas tax increase and new flat-rate mileage tax compares with support for variants on these options. (Appendix A of (5) presents the questionnaire and complete results of the survey.) Survey Respondents The survey respondents were generally representative of the U.S. adult population in terms of Census region and socio-demographic characteristics, although the sample diverged from the national average more (from 5 to 12 percentage points) along a few dimensions of ethnicity, race, education, and age (see Table 1 of (5)). Overall Support Levels for the Transportation Tax Options The survey results show that a majority of Americans would support higher taxes for transportation under certain conditions (see Figure 1). For example, a gas tax increase of 10 per gallon to improve road maintenance was supported by 69% of respondents, whereas support levels dropped to 25% if the revenues were to be used more generally to maintain and improve the transportation system. The only other variant on a gas tax that received at least 60% support

7 Agrawal & Nixon 7 in 2014 was an increase of 10 per gallon with the revenues dedicated to reducing accidents and improving safety. However, support for several other tax options was still above 50%, a healthy showing of support for a tax increase of any kind. For tax options where the revenues were to be spent for undefined transportation purposes, support levels varied considerably by what kind of tax would be imposed, with a new national sales tax roughly twice as popular as either the 10 per gallon gas tax increase a or new mileage tax with a flat rate of 1 per mile.

8 Agrawal & Nixon 8 spent to maintain streets, roads, and highways 69 spent on projects to reduce accidents and improve safety 63 spent to add more modern, technologically advanced systems 57 spent to reduce local air pollution 54 spent to reduce global warming sales tax 49 Mileage tax: rate varies by vehicle s pollution level (average 1 per mile) 43 Gas tax: 10 increase with information on average driver's annual cost 42 Gas tax: 2 increase 41 Gas tax: 10 increase 25 Mileage tax: flat rate of 1 per mile Respondent support (%) FIGURE 1 Support for the Tax Options Surveyed in (Note: Support is the sum of those who said they strongly or somewhat supported the tax option.)

9 Agrawal & Nixon 9 Support by Population Subgroups We also examined support levels for the different tax options by subgroups within the population. The statistical test of two proportions was used to check whether differences among subgroups (e.g., men versus women) are statistically significant at the 95% and 99% confidence levels. This section of the paper summarizes the key findings from this analysis. Readers interested in reviewing the detailed findings should see Tables 2 through 5 in (5). The following discussion focuses on those differences among subgroups where the patterns are clearest. A pattern is defined as clear when (1) the variation in support is statistically significant across at least five of the tax options, and (2) the average magnitude of the difference between the groups across all 11 tax options is at least 7 percentage points or more. Readers should note that the variations described below are not necessarily the only important ones that may exist. Rather, the variations discussed are those that could be identified by the particular statistical tests used and also fell within the cutoff points selected. The clear socio-demographic patterns that emerge are linked to race, employment status, and age. With respect to race, Asians/Asian-Americans were, on average, 17 percentage points more likely to support each tax than whites. Similarly, African-Americans were on average 10 percentage points more likely to support each tax than were whites. As for age, respondents in the youngest group (18- to 24-year olds) were more likely to support virtually all of the taxes than respondents in the two older groups, especially as compared to the oldest group (55 years and older). The average difference in support for the taxes was 17 percentage points for the youngest group as compared to the oldest group. Finally, employed respondents were more supportive of the taxes than retirees, mirroring the differences in support by age. Political party affiliation played a strong role, with support for all of the taxes more likely among registered Democrats than among registered Republicans, voters registered with other parties, or registered voters who are party-independent. The level of support differed for registered Democrats and registered Republicans by an average of 19 percentage points across the 11 tax options. In addition, people who were not registered to vote were modestly more likely to support most of the taxes than were registered voters. The survey asked three questions about travel behavior and personal vehicle mileage in order to examine whether support for the tax options varied by these factors. Respondents who reported driving from 1 to 7,500 miles annually were more likely to support the taxes than people who reported driving more than 12,500 miles annually, but they were less likely to support the taxes than people who said they did not drive at all. Also, respondents who said that they had taken public transit within the previous 30 days were more likely to support the tax options than respondents who said that they had not. Another set of analyses examined how support for the different tax options correlates with respondents opinions about the transportation system. Respondents opinions about road and transit services in their local community are not clearly correlated with support for the taxes, but the quality of local public transit is. Respondents who rated the quality of public transit service in their community as very good were more likely to support the taxes than those who said they had no public transit service at all in their communities. Another set of questions asked respondents about their priorities for how governments might spend transportation revenues: reducing traffic congestion; maintaining streets, roads, and highways; expanding and improving local public transit service; reducing accidents and improving safety; and increasing the use of modern technologies. Not surprisingly, respondents who placed a high priority on these goals were more likely to support almost every tax option than were those who assigned them a low

10 Agrawal & Nixon 10 priority. The differences were, on average, very large over 20 percentage points in three cases. In addition, respondents who placed a high priority on the goals of improving public transit service and increasing the use of modern technology were more supportive of the taxes than those who had assigned those goals a medium priority. Support for Different Versions of the Mileage and Gas Taxes A central goal of the survey was to test how public support varied for different mileage and gas tax proposals. In this study, the base-case proposals for each type of tax were the flat-rate mileage tax of 1 per mile and the 10 gas tax increase without any additional detail. For comparative purposes, respondents were also asked about a single variant of the mileage tax (a variable tax based on how much pollution a vehicle produces) and a series of variants on the gas tax (several proposals that dedicate additional revenues to specific purposes, a phased-in tax increase, and a proposal that informs respondents of the typical annual cost). Figure 2 shows how variants on the tax proposals increased support in comparison to the base-case tax options. For both tax types, the base-case version had the lowest support level, and applying the test of two proportions confirmed that in all cases the increase in support is statistically significant.

11 Agrawal & Nixon 11 spent to maintain streets, roads, and highways 44 spent on projects to reduce accidents and improve safety 38 spent to add more modern, technologically advanced systems 32 spent to reduce local air pollution 28 spent to reduce global warming 26 Gas tax: 10 increase with information on average driver's annual cost 17 Gas tax: 2 increase per year for 5 years 16 Mileage tax: rate varies by vehicle s pollution level (average 1 per mile) Percentage-point increase in support for the variant compared with the base-case tax option FIGURE 2 Relative Increases in 2014 Support for Variations of the Base-Case Gas Tax and Mileage Tax Concepts. (Note: Support is the sum of those who said they strongly or somewhat supported the tax option.) We also looked at the change in support levels for each tax variant by respondent subgroups that are defined by Census region, socio-demographic and political characteristics, travel behavior characteristics, and opinions about the transportation system. Overall, the analysis considered 63 population subgroups, for each of which there are 8 tax comparisons, resulting in a total of 504 cases examined. (Readers interested in reviewing the detailed findings should see Tables 6 through 9 in (5).)

12 Agrawal & Nixon 12 The overall pattern of increased support for the variants holds for the subgroups, just as for the respondent pool as a whole. Across all 504 cases examined, the tax variants improved support in all but one case (and this one case was for a subgroup with fewer than 50 respondents). The increase in support for the variants as compared to the base cases was statistically significant for 93% of cases. Further, the increases were very large: At least 10 percentage points for 97% of cases At least 20 percentage points for 74% of cases At least 30 percentage points for 42% of cases At least 40 percentage points for 17% of cases Comparison of Support over Time ( ) Most of the survey questions replicate those in the four surveys previously administered in this series, so it is possible to look at trends in support over time. The trend analysis shows that support levels have changed only a little over the five surveys (see Figure 3). In most cases the support for a tax varied by 5 or fewer percentage points from 2010 to 2011 to 2012, a change too small to suggest a meaningful change in support. However, Americans were modestly more willing to support most of the tax increases in 2013 and 2014 than they were in the previous three years. In 2014, support levels were at their highest ever for 7 of the tax options tested. The one striking exception to the trend of fairly similar support levels across all the taxes for all five surveys is the gas tax increase with revenues dedicated to projects that reduce air pollution. Here, support has varied considerably from year to year, with a low of 30% support in 2010 to a high of 54% support in 2014.

13 Agrawal & Nixon 13 spent to maintain streets, roads, and highways spent on projects to reduce accidents and improve safety spent to add more modern, technologically advanced systems spent to reduce local air pollution spent to reduce global warming 0.5 sales tax Mileage tax: rate varies by vehicle s pollution level (average 1 per mile) Gas tax: 10 increase with information on average driver's annual cost Gas tax: 2 increase per year for 5 years Gas tax: 10 increase Mileage tax: flat rate of 1 per mile Respondent support (%) FIGURE 3 Trends in Support for the Tax Options, (Note: Support is the sum of those who said they strongly or somewhat supported the tax option.)

14 Agrawal & Nixon 14 We also found that a few population subgroups were clearly more likely to support the taxes across all five surveys: Asians/Asian-Americans and blacks/african-americans (compared to whites) Younger people (compared to people in both older age groups) Democrats (compared to Republicans and Independents) People who drove the fewest miles per year (compared to people who did not know how many miles per year they drove or who did not drive) People who had used transit in the previous 30 days (compared to people who did not) People who think government should place a high priority on expanding and improving local public transit service, maintaining streets and roads, reducing accidents and improving safety, and using modern technology (compared with people who think government should place a low priority on these goals). Our analysis of how the tax variations boosted support over the base cases shows relatively little change from 2010 to 2014 (Figure 4). In every case, the variations had higher support levels than the base-case options, and the boosts in support were fairly similar each year the questions were asked. The gas tax variant with the most consistent boost has been the option to phase in the tax increase over five years. The increase in support for this variant has remained within a 4 percentage point range across all survey years. The boosts for the other gas tax variants have fluctuated a little more but still within a range of 9 points or fewer, except for the gas tax linked to projects that would reduce local air pollution. This variant received a small boost in support in 2010 but then received relatively large boosts since then (between 21 and 30 percentage points). Additionally, there has been a gradual but steady increase in support for the mileage tax with variable rates based on vehicle emissions. For that tax variant, the boost has climbed from 12 percentage points in 2010 to 25 percentage points in 2014.

15 Agrawal & Nixon 15 spent to maintain streets, roads, and highways spent on projects to reduce accidents and improve safety spent to add more modern, technologically advanced systems spent to reduce local air pollution spent to reduce global warming Gas tax: 10 increase with information on average driver's annual cost Gas tax: 2 increase per year for 5 years Mileage tax: rate varies by vehicle s pollution level (average 1 per mile) Percentage-point increase in support for the variant compared with the base-case tax option FIGURE 4 Changes over Time for the Relative Increases in Support for Variations of the Base- Case Gas Tax and Mileage Tax Concepts ( ). (Notes: Support is the sum of those who said they strongly or somewhat supported the tax option. The base-case proposals were a new flat-rate mileage tax of 1 per mile and a 10 per gallon gas tax increase, without any additional detail)

16 Agrawal & Nixon 16 FINDINGS RELATED TO PUBLIC TRANSIT For 2012, 2013, and 2014 a new emphasis in the survey project was to understand various perceptions related to public transit, including knowledge and opinions about federal taxes to support transit. This section pulls together the different pieces of the survey to highlight all findings related to transit. A question early in the survey asked respondents their opinion on the quality of public transit in their community. The majority of respondents (57%) said that it is very or somewhat good, 15% said that it is poor, and 28% said either that there is no service in their community or that they do not know about transit quality. These values are very close to those from identical questions asked in all four prior surveys. Another early series of questions in the survey asked respondents how highly they would prioritize various things government could do to improve the transportation system for everyone in the state where you live. One of the priorities tested was expanding and improving local public transit service. Public transit was a high priority for close to half of respondents (44%), though this was the lowest percentage among the five priorities tested. However, when looking at those who felt transit was either a high or medium priority, transit rated not so differently from the other options 79% of respondents felt this way, compared to the other options that ranged from a low of 80% to a high of 95%. The two most popular priorities were road maintenance and improving safety. Later in the survey, respondents were asked if they knew how the cost of providing transit service is covered. The first question in the series was as follows: When people ride public transit, they pay a fare. This money is used to pay for the service. Do you think that the money collected from public transit fares in general covers the full cost of the service? Thirtyfour percent of respondents (incorrectly) said yes, 16% said that they did not know, and only 50% (correctly) said no. These responses are similar to those from the 2013 survey. Those respondents who did not think that fares cover the full costs of transit were asked some follow-up questions. First, they were asked, In general, what percent of the full cost of public transit services do you think the fares cover? Twenty-eight percent said that fares cover 1 to 33% of the full cost, 38% said that fares cover 34 to 66% of the full cost, 17% said that fares cover 67 to 100% of the full cost, and 18% said that they did not know. For those respondents who did not think that fares cover all transit costs, the survey asked if they thought the federal, state, and local government also helps to pay for public transit services around the country. Slightly more than half (55%) knew that the federal government helps pay for transit, with more respondents aware of the local contribution (64%) and the state contribution (76%). An alternative way to think about the survey findings on this topic is in terms of the percent of all respondents who were aware of the role each government entity plays in funding. Calculating the numbers this way, 36%of all respondents knew the federal government pays for transit, 42% knew of the local government role, and 50% knew of the state government role. Finally, a set of questions delved into respondents beliefs about the best ways for Congress to help pay for transit. The first of these asked respondents the following question: Now I have a question about whether or not GAS tax money should be spent to pay for public transit. Some people say that money from gas taxes should only be spent on roads and highways, since drivers pay the tax. Other people say gas tax money should be used to pay for public transit IN ADDITION to roads and highways, because transit helps reduce traffic congestion and wear-and-tear on the roads. Would you support or oppose spending SOME gas tax money on public transit?

17 Agrawal & Nixon 17 Sixty-four percent of respondents supported spending gas tax revenues on transit and 36% opposed this. A multipart question then posed the scenario that Congress had decided to spend more money on public transit but had not decided how to pay for this. Respondents were first asked whether they would support each of the following three options to pay for expanding and improving public transportation: reducing spending on other federal programs, raising transit fares, or raising the federal gas tax. In 2014, reducing federal spending on other programs received the most support (60%), followed by raising transit fares (52%), and trailed by raising the federal gas tax (36%). When respondents were asked which of the three choices they preferred, the same hierarchy emerged: 48% preferred reducing spending on other programs, 24% preferred raising transit fares, and 17% preferred raising the federal gas tax. Across the three years of surveying, there was a statistically significant increase in support for each individual option from 2012 to 2014: four percentage points more support for reducing federal spending, seven points more support for raising transit fares, and eight points more support for raising the federal gas tax. However, the percent of respondents choosing each option as their preferred alternative remained almost the same from year to year. SUMMARY OF KEY FINDINGS Support Levels for the 11 Tax Options When interpreting the survey results, it is important to keep in mind that the questionnaire described the various tax proposals in only general terms, so the results cannot be assumed to reflect support for any actual proposal put forward. Nevertheless, the results show likely patterns of support and, more important, the public s relative preferences among different transportation tax options. With this caveat in mind, the survey results show that a majority of Americans would support higher taxes for transportation under certain conditions. For example, a gas tax increase of 10 per gallon to improve road maintenance was supported by 69% of respondents, whereas support levels dropped to 51% if the revenues were to be devoted to reducing global warming, or only 25% if the revenues were to support undefined transportation purposes. As for tax options where the revenues were to be spent for undefined transportation purposes, support levels varied considerably by the kind of tax that would be imposed, with a sales tax much more popular (49%) than either a gas tax increase (25%) or a new mileage tax (19%). A central goal of the survey was to compare public support for two alternative versions of the mileage tax and eight versions of a gas tax increase. Variations on the two taxes increased support substantially over that for the base case of each (a flat-rate mileage tax of 1 per mile and a 10 gas tax increase proposed without any additional detail). Those boosts in support ranged from a low of 16 percentage points to a high of 44 points. In addition to examining support for the different tax options among the overall population, we examined support by subgroups within the population. Breaking the population into subgroups by socio-demographic categories reveals surprisingly few links with support for the taxes. Some of the only statistically significant patterns that emerge are that the taxes generally had greater support from respondents who were younger, non-white, Democrat, not registered to vote, did not drive at all within the past year or did not know their mileage, drove the least (1 to 7,500 miles per year), rated transit service in their community as very good, or

18 Agrawal & Nixon 18 placed a high priority on having government improve various aspects of the transportation system in their state. When comparing support by population subgroup for the gas tax and mileage tax variations to the base-case versions, the overall picture that emerges is simple and clear: the base-case taxes were less popular than the alternative tax options for virtually every subgroup. Further, that boost in support for the variants is generally quite large. We examined 504 cases (8 tax variants for each of 63 subgroups) and found that the boost in support for the variant was at least 30 percentage points for 42% of the cases. Changes in Support for the 11 Tax Options, Our surveys indicate that American public opinion about the federal transportation tax options tested has changed very little since The 2014 survey found approximately the same support for the tax increases as in the previous four years, though support levels have generally risen a bit over the five-year period and were the highest ever in 2014 for seven of the tax options. In addition, the analysis of how the variations on the gas and mileage taxes boosted support over the base cases for each shows very little change from one year to the next. The only truly notable change over time is the continued rise in support for a variable-rate mileage tax based on vehicle pollution levels. The boost in support for this tax option has risen from 12 percentage points in 2010 to 25 points in The fact that all five surveys generally show such similar results suggests that the views expressed are indeed generally representative of the American public and are not aberrations caused by an unusual and unrepresentative sample in any year of the survey. Knowledge and Preferences Related to Public Transit The questions focused on public transit revealed that a very high percentage of people (79%) placed a high or medium priority on improving and expanding public transit in their state, though some other priorities had even higher support levels. Many respondents were not aware of the different government entities that fund transit. Knowledge was the lowest about the federal role; only 36% of people knew that the federal government helps to pay for public transit. As to how respondents wanted to see the federal government pay for improving and expanding public transit, raising neither the gas tax nor transit fares was particularly popular. The most popular option was to cut spending on other government programs. IMPLICATIONS FOR PRACTICE The results of the five surveys suggest several implications for transportation professionals and policymakers who wish to craft transportation revenue increases that will be more appealing or at least less objectionable to the public: Careful program design can increase support for higher gas taxes or a new mileage tax. The survey results show that the very low support levels for a gas tax increase or a new mileage tax can be raised by modifying how the tax is structured and the way it is described. For example, support rises when revenues are dedicated to specific purposes popular with the public, the tax increase is spread out over several years, or information is provided about how much the increase will cost drivers annually.

19 Agrawal & Nixon 19 Stressing the environmental, safety, and maintenance benefits will increase support for transportation taxes, including ones for transit. Devoting revenues to maintenance and safety can increase support levels substantially across the whole population. Also, linking a transportation tax to environmental benefits can strongly increase support among most population subgroups. Linking transit with environmental benefits may be a particularly successful way to increase support for transit revenues. Demographic change in the U.S. population may increase support for transportation taxes. The surveys found that the youngest respondents were much more supportive of the tax options than older respondents. If this variation reflects a true generational shift, then these opinions would persist as those currently young respondents age and might also hold with the age cohorts behind them who soon become adults. FUTURE RESEARCH NEEDS It should be noted that the results presented in this paper are derived from bivariate analysis only. In order to more fully explore the complex reasons behind individuals support or opposition to the transportation taxes surveyed, the authors future research plans include conducting multivariate analyses that pool all five years of survey data. ACKNOWLEDGEMENTS The authors thank the Mineta Transportation Institute at San José State University for funding this research and the institute s staff for their support, especially Research Director Karen Philbrick. In addition, we thank Stephen Granger-Bevan, who provided research and data analysis assistance, Laura Gil-Trejo, Director of the Social Science Research Center at California State University, Fullerton, who managed the survey implementation, and the 1,503 individuals who completed the survey. The views expressed are the authors alone, as is the responsibility for any errors or omissions.

20 Agrawal & Nixon 20 REFERENCES 1. Agrawal, Asha Weinstein, and Hilary Nixon. What Do Americans Think About Federal Transportation Tax Options? Results from a National Survey. San Jose, CA: Mineta Transportation Institute, June %20Annual%20Trans.%20Survey%20( ).pdf. Accessed July 21, Agrawal, Asha Weinstein, and Hilary Nixon. What Do Americans Think About Federal Transportation Tax Options? Results From Year 2 of a National Survey. San Jose, CA: Mineta Transportation Institute, June Agrawal, Asha Weinstein, Hilary Nixon, and Vinay Murthy. What Do Americans Think About Federal Tax Options to Support Public Transit, Highways, and Local Streets and Roads? Results from Year 3 of a National Survey. San Jose, CA: Mineta Transportation Institute, June Agrawal, Asha Weinstein, and Hilary Nixon. What Do Americans Think About Federal Tax Options to Support Public Transit, Highways, and Local Streets and Roads? Results from Year 4 of a National Survey. San Jose, CA: Mineta Transportation Institute, June Agrawal, Asha Weinstein, and Hilary Nixon. What Do Americans Think About Federal Tax Options to Support Public Transit, Highways, and Local Streets and Roads? Results from Year 5 of a National Survey. San Jose, CA: Mineta Transportation Institute, June Pew Research Center for the People and the Press. Assessing the Representativeness of Public Opinion Surveys. Washington, DC: Pew Research Center for the People and the Press, May 15, AAPOR Cell Phone Task Force. New Considerations for Survey Researchers When Planning and Conducting RDD Telephone Surveys in the U.S. with Respondents Reached Via Cell Phone Numbers. Deerfield, IL: American Association of Public Opinion Research, ntdisplay.cfm&contentid= United States Census Bureau American Community Survey 1-Year Estimates. No date. Downloaded from